[PATCH 07/12] v2: improve deduplication checks

["Eric Wong (Contractor, The Linux Foundation)" <e@80x24.org>]

First off, decode text portions of messages since some archived
mail I got was converted from quoted-printable or base-64 to
8bit by the original recipient.  Attempting to merge them with
my own archives (which had no conversion done) led to
unnecessary duplicates showing up.

Then, normalize CRLF line endings in text portions to LF.

In the headers, we relax the content_id hashing to ignore quotes
and lower-case domain names in To, Cc, and From headers since
some mail processors will alter them.

Finally, I've discovered Email::MIME->new($mime->as_string)
does not always round-trip reliably, so we calculate the
content_id twice on user-supplied messages.
---
 lib/PublicInbox/ContentId.pm  | 67 +++++++++++++++++++++++++++++------
 lib/PublicInbox/V2Writable.pm | 35 +++++++++++++-----
 t/content_id.t                | 10 ++++++
 3 files changed, 94 insertions(+), 18 deletions(-)

diff --git a/lib/PublicInbox/ContentId.pm b/lib/PublicInbox/ContentId.pm
index 279eec0..b1d27eb 100644
--- a/lib/PublicInbox/ContentId.pm
+++ b/lib/PublicInbox/ContentId.pm
@@ -7,10 +7,19 @@ use warnings;
 use base qw/Exporter/;
 our @EXPORT_OK = qw/content_id content_digest/;
 use PublicInbox::MID qw(mids references);
+use PublicInbox::MsgIter;
 
 # not sure if less-widely supported hash families are worth bothering with
 use Digest::SHA;
 
+sub digest_addr ($$$) {
+	my ($dig, $h, $v) = @_;
+	$v =~ tr/"//d;
+	$v =~ s/@([a-z0-9\_\.\-\(\)]*([A-Z])\S*)/'@'.lc($1)/ge;
+	utf8::encode($v);
+	$dig->add("$h\0$v\0");
+}
+
 sub content_digest ($) {
 	my ($mime) = @_;
 	my $dig = Digest::SHA->new(256);
@@ -27,24 +36,62 @@ sub content_digest ($) {
 	}
 	foreach my $mid (@{references($hdr)}) {
 		next if $seen{$mid};
-		$dig->add('ref: '.$mid);
+		$dig->add("ref\0$mid\0");
 	}
 
 	# Only use Sender: if From is not present
 	foreach my $h (qw(From Sender)) {
-		my @v = $hdr->header_raw($h);
+		my @v = $hdr->header($h);
 		if (@v) {
-			$dig->add("$h: $_") foreach @v;
-			last;
+			digest_addr($dig, $h, $_) foreach @v;
 		}
 	}
-
-	# Content-* headers are often no-ops, so maybe we don't need them
-	foreach my $h (qw(Subject Date To Cc)) {
-		my @v = $hdr->header_raw($h);
-		$dig->add("$h: $_") foreach @v;
+	foreach my $h (qw(Subject Date)) {
+		my @v = $hdr->header($h);
+		foreach my $v (@v) {
+			utf8::encode($v);
+			$dig->add("$h\0$v\0");
+		}
+	}
+	# Some mail processors will add " to unquoted names that were
+	# not in the original message.  For the purposes of deduplication,
+	# do not take it into account:
+	foreach my $h (qw(To Cc)) {
+		my @v = $hdr->header($h);
+		digest_addr($dig, $h, $_) foreach @v;
 	}
-	$dig->add($mime->body_raw);
+	msg_iter($mime, sub {
+		my ($part, $depth, @idx) = @{$_[0]};
+		$dig->add("\0$depth:".join('.', @idx)."\0");
+		my $fn = $part->filename;
+		if (defined $fn) {
+			utf8::encode($fn);
+			$dig->add("fn\0$fn\0");
+		}
+		my @d = $part->header('Content-Description');
+		foreach my $d (@d) {
+			utf8::encode($d);
+			$dig->add("d\0$d\0");
+		}
+		$dig->add("b\0");
+		my $ct = $part->content_type || 'text/plain';
+		my $s = eval { $part->body_str };
+		if ($@ && $ct =~ m!\btext/plain\b!i) {
+			# Try to assume UTF-8 because Alpine
+			# seems to do wacky things and set
+			# charset=X-UNKNOWN
+			$part->charset_set('UTF-8');
+			$s = eval { $part->body_str };
+		}
+		if (defined $s) {
+			$s =~ s/\r\n/\n/gs;
+			$s =~ s/\s*\z//s;
+			utf8::encode($s);
+		} else {
+			$s = $part->body;
+		}
+		$dig->add($s);
+	});
 	$dig;
 }
 
diff --git a/lib/PublicInbox/V2Writable.pm b/lib/PublicInbox/V2Writable.pm
index 0dcdeda..e9fd502 100644
--- a/lib/PublicInbox/V2Writable.pm
+++ b/lib/PublicInbox/V2Writable.pm
@@ -259,12 +259,32 @@ sub purge_oids {
 	$purges;
 }
 
+sub content_ids ($) {
+	my ($mime) = @_;
+	my @cids = ( content_id($mime) );
+
+	# Email::MIME->as_string doesn't always round-trip, so we may
+	# use a second content_id
+	my $rt = content_id(PublicInbox::MIME->new(\($mime->as_string)));
+	push @cids, $rt if $cids[0] ne $rt;
+	\@cids;
+}
+
+sub content_matches ($$) {
+	my ($cids, $existing) = @_;
+	my $cid = content_id($existing);
+	foreach (@$cids) {
+		return 1 if $_ eq $cid
+	}
+	0
+}
+
 sub remove_internal {
 	my ($self, $mime, $cmt_msg, $purge) = @_;
 	$self->idx_init;
 	my $im = $self->importer unless $purge;
 	my $over = $self->{over};
-	my $cid = content_id($mime);
+	my $cids = content_ids($mime);
 	my $parts = $self->{idx_parts};
 	my $mm = $self->{mm};
 	my $removed;
@@ -287,7 +307,7 @@ sub remove_internal {
 			}
 			my $orig = $$msg;
 			my $cur = PublicInbox::MIME->new($msg);
-			if (content_id($cur) eq $cid) {
+			if (content_matches($cids, $cur)) {
 				$smsg->{mime} = $cur;
 				$gone{$smsg->{num}} = [ $smsg, \$orig ];
 			}
@@ -572,8 +592,7 @@ sub get_blob ($$) {
 sub lookup_content {
 	my ($self, $mime, $mid) = @_;
 	my $over = $self->{over};
-	my $cid = content_id($mime);
-	my $found;
+	my $cids = content_ids($mime);
 	my ($id, $prev);
 	while (my $smsg = $over->next_by_mid($mid, \$id, \$prev)) {
 		my $msg = get_blob($self, $smsg);
@@ -582,16 +601,16 @@ sub lookup_content {
 			next;
 		}
 		my $cur = PublicInbox::MIME->new($msg);
-		if (content_id($cur) eq $cid) {
+		if (content_matches($cids, $cur)) {
 			$smsg->{mime} = $cur;
-			$found = $smsg;
-			last;
+			return $smsg;
 		}
 
+
 		# XXX DEBUG_DIFF is experimental and may be removed
 		diff($mid, $cur, $mime) if $ENV{DEBUG_DIFF};
 	}
-	$found;
+	undef;
 }
 
 sub atfork_child {
diff --git a/t/content_id.t b/t/content_id.t
index adcdb6c..01ce65e 100644
--- a/t/content_id.t
+++ b/t/content_id.t
@@ -22,4 +22,14 @@ my $orig = content_id($mime);
 my $reload = content_id(Email::MIME->new($mime->as_string));
 is($orig, $reload, 'content_id matches after serialization');
 
+foreach my $h (qw(From To Cc)) {
+	my $n = '"Quoted N\'Ame" <foo@EXAMPLE.com>';
+	$mime->header_str_set($h, "$n");
+	my $q = content_id($mime);
+	is($n, $mime->header($h), "content_id does not mutate $h:");
+	$mime->header_str_set($h, 'Quoted N\'Ame <foo@example.com>');
+	my $nq = content_id($mime);
+	is($nq, $q, "quotes ignored in $h:");
+}
+
 done_testing();
-- 
EW